This program has led to findings that shed light on: 1) the remarkable sensitivity of family processes to the individual psychological development of parents and adolescents; 2) variations in the patterns of maturation of adolescents, distinguished by ego development trajectories; 3) transformations in adolescent-parent relationships through middle adolescence; and, 4) the role of the family and individual development in regulating key aspects of a chronic medical illness in adolescence, diabetes. These 4 insights lead directly to my specific aims, implemented through the 3 proposed component projects: I. Family Contexts of Diverse Paths of Adolescent Development, a set of further studies of ego development trajectories, focusing on unraveling linkages between adolescent ego development and family processes; II. Adolescent and Family Antecedents of Young Adult Social Development, to discover how family and individual forces in adolescence influence social development (attachments, the nature of intimate relationships, antisocial behavior) in early adult years; III. Family Contexts of Outcomes of Diabetic Adolescents. Building on the substantive and methods advances of the other 2 projects, this component analyzes how an adolescent's new-onset chronic medical illness intersects with development and family processes. It is organized around 2 broad questions: 1) How do individual and family psychosocial dimensions predict diabetes-specific outcomes (diabetes adjustment, compliance, metabolic control, and early complications)? and, 2) How does the onset of diabetes in adolescence influence the development of individual patients and their families? All 3 of these components are based on longitudinal data. The research design for the young and diabetic components involve obtaining a new wave of theoretically guided individual assessments, at one time point. Methods include interview, self reporting, peer ratings (Q-sort), and precise biological assessments (hemoglobin A1, Fundoscopic measurements, microalbumin). Statistical analyzes are primarily multivariate approaches, involving MANCOVA, hierarchical multiple regressions, and multiple regressions, and multiple discriminant functions analyses where appropriate.